Modelpublic

Rorch/test_pelby_v2

sourceHugging Faceapache-2.0updated 2y agoView on Hugging Face

0likes7downloads

Model Card

wav2vec2-full_v2

This model is a fine-tuned version of Osolon/wav2vec2-large-xls-r-300m-pl on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: inf
—Wer: 0.2842

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

—learning_rate: 0.0001
—trainbatchsize: 8
—evalbatchsize: 4
—seed: 42
—gradientaccumulationsteps: 2
—totaltrainbatch_size: 16
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—lrschedulerwarmup_steps: 100
—num_epochs: 4
—mixedprecisiontraining: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.2498	0.1241	400	inf	0.2197
0.2049	0.2481	800	inf	0.1586
0.1798	0.3722	1200	inf	0.1380
0.1527	0.4963	1600	inf	0.1344
0.1588	0.6203	2000	inf	0.1336
0.1401	0.7444	2400	inf	0.1168
0.1336	0.8685	2800	inf	0.1064
0.1308	0.9926	3200	inf	0.0983
0.1177	1.1166	3600	inf	0.0911
0.1261	1.2407	4000	inf	0.0878
0.2208	1.3648	4400	inf	0.1019
0.2814	1.4888	4800	inf	0.1836
0.4642	1.6129	5200	inf	0.2173
0.6645	1.7370	5600	inf	0.6257
0.8667	1.8610	6000	inf	0.7893
0.7724	1.9851	6400	inf	0.7634
0.6747	2.1092	6800	inf	0.5521
0.5989	2.2333	7200	inf	0.5029
0.5234	2.3573	7600	inf	0.4523
0.4844	2.4814	8000	inf	0.3895
0.4708	2.6055	8400	inf	0.3319
0.4701	2.7295	8800	inf	0.2781
0.4665	2.8536	9200	inf	0.2713
0.4652	2.9777	9600	inf	0.2844
0.4539	3.1017	10000	inf	0.2842
0.4692	3.2258	10400	inf	0.2842
0.4658	3.3499	10800	inf	0.2842
0.4602	3.4739	11200	inf	0.2842
0.4648	3.5980	11600	inf	0.2842
0.4608	3.7221	12000	inf	0.2842
0.4677	3.8462	12400	inf	0.2842
0.4731	3.9702	12800	inf	0.2842

Framework versions

—Transformers 4.44.2
—Pytorch 2.4.1+cu121
—Datasets 2.21.0
—Tokenizers 0.19.1